STUDIES OF COMPOUND STATES OF NEGATIVE * IONS USING LASER BEAMS* I Office of Naval Research Task NR 393-071 R. N. Compton and G. D. Alton Oak Ridge National Laboratory Oak Ridge, Tennessee 37831 August 1985 ~B * Research sponsored by the Office of Naval Research under Interagency SAgreement Number 40-1147-81. Oak Ridge National Laboratory is operated 1= by Martin Marietta Energy Systems, Inc., with the U.S. Department of 7 *T Energy under contract DE-AC05-840R2 1400. CLOT~ve be-, its ~swee \ 'Rc
55
Embed
STUDIES OF COMPOUND STATES OF NEGATIVE * IONS USING LASER BEAMS* · 2015. 1. 15. · STUDIES OF COMPOUND STATES OF NEGATIVE * IONS USING LASER BEAMS* I Office of Naval Research Task
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
STUDIES OF COMPOUND STATES OF NEGATIVE
* IONS USING LASER BEAMS*
I Office of Naval Research Task NR 393-071
R. N. Compton and G. D. Alton
Oak Ridge National Laboratory
Oak Ridge, Tennessee 37831
August 1985 ~B
* Research sponsored by the Office of Naval Research under Interagency
SAgreement Number 40-1147-81. Oak Ridge National Laboratory is operated
1= by Martin Marietta Energy Systems, Inc., with the U.S. Department of
6a NAME OF PERFORMING ORGANIZATION 6b OFFICE SYMBOL 7a NAME OF MONITORING ORGANIZATION
Oak Ridge National Laborat ry (If applicable) N/A
6c. ADDRESS (City, State, and ZIPCode) 7b ADDRESS (City, State, and ZIP Code)P.O. Box X N/AOak Ridge, Tennessee 37831
8a NAME OF FUNDING Y SPONSORING 1b OFFICE SYMBOL 9 PROCUREMENT INSTRUMENT iDENTIFiCATION NuMBERORGANIZATION (If applicable)Office of Naval Research N00014-85-F-0011
8c. ADDRESS (City, State, and ZIP Code) 10 SOURCE OF FUNDING NUMBERSPhysics Division, Code 421 PROGRAM PROJECT ASK 'jWORK UNIT
800 North Quincy Street ELEMENT NO NO NO ACCESSION NO
Arlington, Virginia 22217 61153N IRR 4011-03 1 NR393-07111 TITLE (Include Security Classification)
Studies of Compound States of Negative Ions Using Laser Beams
12 PERSONAL AUTHOR(S)R. N. Compton and G. D. Alton
13a TYPE OF REPORT 13b TIME COVERED 14 DATE OF REPORT (Year, Month, Day) 15 PAGE COUNTSummary FROM 1L/12/84TO9/30/85 August 15, 1985
!6 SUPPLEMENTARY NOTATION Prepared in cooperation with Dr. D. J. Pegg and J. S. Thompson (Universi yof Tennessee), T. J. Kvale (ORAU Postdoctoral Fellow) A. Dodhy (Graduate Student, Auburn)
17 COSATI CODES 8 SUBJECT TERMS (Continue on reverse if necessary and identify by block number)FIELD GROUP SUB-GROUP [Laser photodetachment of.negative ion beams. Neutralization
of negative ion beams Autodetachment spectroscopy ofnegative ion beams; Electron collisions in strong laser field
19 ABSTRACT (Continue on reverse if necessary and identify by block number)This research involves experimental studies of compound states of negative ions using laserbeams. Negative ion beams of stable (H-C, 0 , etc.) and metastable (He-, Be_' Ca-, ?tc.)species are generated by double charge exchange from positive ion beams colliding with alkalivapor. Measurements of the photodetachment and autodetachment spectra for these ions providenew information on compound state energy levels. In addition experiments are underway tostudy the collisions of electrons with atoms in strong laser fields. Many theoreticalpredictions have been presented recently on this subject. These experiments should providethe first tests of such predictions.,;
20 DISTRIBUTION/ AVAILABILITY OF ABSTRACT 21 ABSTRACT SECURITY CLASSIFICATION
UNCLASSIFIED/UNLIMITED 0 SAME AS RPT Q1 DTIC USERS22a NAME OF RESPONSIBLE INDIVIDUAL 221t N1 ffqrde Area Code) 22c OFFICE SYMBOLR. N. Compton If -
DO FORM 1473,84 MAR 83 APR edition may be used until exhausted. SECURITY CLASSIFICATION OF THIS PAGEAll other editions are obsolete
STUDIES OF COMPOUND STATES OF NEGATIVE IONS
USING LASER BEAMS
R. N. Compton and G. D. Alton
Oak Ridge National Laboratory, Oak Ridge, Tennessee 37831
This is a second-year progress report and third-year proposal on our research effort
entitled "Studies of Compound States of Negative Ions Using Laser Beams" under
Office of Naval Research (ONR) task NR 393-071. The reader will note a change in
the title of the proposal from that of the previous year, namely, "... Atomic Negative
Ions" is replaced by "... Negative Ions" in order to reflect the new work we will
describe on molecular negative ions (He2-). Our proposed and completed research
• effort can be separated into two well-defined activities: (I) Autodetachment,
Collisional Detachment, and Photodetachment Studies of H-, He-, Be-, Ca-, and
He- and (II) Experimental Studies of Electron Collisions in the Field of a Laser.
Two Physical Review and one Physical Review Letters articles have been published.
One article describing studies of Ca production has been accepted by Nuclear
Instrumentation and Methods, and Physical Review Letters has accepted our
manuscript entitled "Experimental Determination of the Energy Level of Be
(Is22s2p 2)4P." A third Physical Review Letters article entitled "Autodetachment
Spectroscopy of He2-" is in final preparation and will be submitted soon. All of these
articles are included as Appendix I.
The Office of Naval Research support has allowed the researchers to develop this
program into an important basic research component at the Oak Ridge National
Laboratory. The group consists of R. N. Compton (Health and Safety Research
Division) and G. D. Alton (Physics Division), Oak Ridge National Laboratory, D. J.
Pegg, professor of physics, The University of Tennessee, and most importantly, T. J.
Kvale, postdoctoral fellow from The University of Missouri-Rolla. Dr. Pegg has
separate support from the Department of Energy (Basic Energy Sciences) which also
*supports J. S. Thompson, a full-time graduate student from The University of
Tennessee. The Negative Ion Source Test Facility has been primarily funded by the
* Physics Division. ONR funds mainly pay for the support of Dr. T. J. Kvale and
some of the day-to-day operational expenses. Dr. Kvale has mastered the
experimental complexities of the Negative Source Ion Test Facility. Three other
graduate students have been associated with the project in various capacities and are
* responsible for the day-to-day operation of the facility and much of the data
acquisition and analysis. Adila Dodhy, a graduate student from Auburn University,
has been studying laser-induced "continuum-continuum" effects under high laser fields.
Her Ph.D. thesis is now complete and she will be joining Dr. H. Walther's group at
the Max Planck Institute in Garching, West Germany in September 1985. Michael
Shea, graduate student in physics at Vanderbilt University, contributed to the
* development of a laser produced, high resolution, high emittance electron beam.
The past year has been very productive; the Negative Ion Source Test Facility and
the laser experiments have operated continuously. Therefore, there are a number of
results to report, and we will describe only final results, referring the reader to the
papers in Appendix I for details.
We will separate discussions of the progress into two parts. Part I will summarize
the studies of negative ion states of Ca, Be, and He2 . Part 11 will describe
results and progress on high laser field effects on electron-atom collisions. In Part III,
we will present our proposed direction of research. The proposal is strongly influenced
by a recent study of He2 to be described in Part I below.
1. Studies of Negative Ions
The new, modified Negative Ion Source Test Facility is now operating routinely.
This major apparatus was described in detail in the previous progress report and to a
* lesser extent in the papers enclosed in Appendix 1. However, further improvements to
S 3
the facility will be made by installation of a recently procured 200 kV, 10 mA highly
regulated power supply. This will give us the capability of producing and studying
* heavy negative ions such as Ca-. Progress during the early portion of this fiscal year
* was inhibited by severe reliability problems with the commercially procured hollow
cathode positive ion source which made spectroscopic measurements of the type
* discussed in this report difficult, if not impossible. The severe nature of the problem
necessitated a complete redesign and development of a new or more reliable source.
The source has been fabricated, tested, and utilized in the Be- and He2 spectroscopic
investigations. The success of these experiments is largely attributable to this
development. This development constitutes a major improvement to the technology of
hollow cathode positive ion sources. The sources will be the subject of a forthcoming
* report which will be submitted for publication in RIS or Nuclear Instrument ations
*and Methods. Operation of the ion source with beryllium also required special
handling procedures and sometimes made progress cumbersome and oftentimes slow,
*however, we have successfully completed the Be studies as described below.
We have completed measurement of the energy level of Be (Is s2s2p ) P. Details
of these measurements are to be found in Appendix I. Be ions were produced in
sequential charge exchange collision between 50- and 60-keV Be +ions and lithium
*vapor. The energy level of Be was determined directly by measuring the
autodetached electron energies leaving the atom in the ground state. Figure 1 shows
one such measurement. The electron energy scale was calibrated using the collisional
- "cusp" peak at '-3.5 eV. There is an approximately three to one compression of the
lab to center-of-mass energy, thus, the electron energy resolution can be seen to be
-50 meV. Our experimental value is compared with four theoretical values in Table
1. This represents the first experimental measurement of the energy level of a
metastable negative ion of the Group 11 metals. Previous to this measurement only
one other metastable energy level, that of He-( 4 P), had been reported.1I
r
4
We have begun measurements on the energy levels for metastable states of other
Group IIA metals, Mg and Ca In this connection, we have determined
production efficiencies for Ca- These results are described in the preprint found in
Appendix I.
We have also completed and reported studies of collisional detachment of H- ions
with various target gases. Again these results are described in the preprint contained
in Appendix I. This paper represents only the second such study of collisionally
detached *cusp" electron spectroscopy for a negative ion. These studies have relevance
to neutral beam technology.
Perhaps the most interesting results to date are the recent measurements on He 2 .
Long-lived metastable negative ion beams of He2 have been recently produced for the
first time by Bae, Coggiola, and Peterson. 2 The electronic structure of He2 has been
subsequently investigated by Michels3 who finds the 4Ig (ls31so2so pfl,,) state of
He 2- to be bound relative to the lowest triplet state He; (a3 Z+ ) by 0.233 eV. These
theoretical studies further show that the 4 Z" (snegative ion state is
unbound relative to He (a3Z+ ) for internuclear separation < 6 A. For internuclear
." separations near the potenital minimum (-I k), the 4Z+ state tracks sufficiently
close to the a3 Z+ state that it was not possible to determine whether this state is
stable. More extensive calculations of this potential are in progress (H. H. Michels,
private communication). We have recently begun a study of electron autodetachment
spectroscopy of metastable He 2 formed by double charge exchange of energetic
(30-65 keV) He+" ions with lithium vapor. Figure 2 shows a typical electron spectrum
covering the range from 0 to 50 eV in the laboratory system. The steep rise at -- 0 eV
is due to secondary electrons from ion and electron scattering with slits, etc. which
defines the path of the ion beam. The peaks at -2 and 33 eV are due to
autodetachment from He2-. The peak at -2 eV can be resolved into two peaks and
will be discussed below. The peak at 33 eV results from autodetachment from
He 2 (gH~),_0 to 2He (IS) + e as illustrated in Fig. 3. Using the vibrational
b5
wavefunctions from the calculations of Michels 3 and the average of a large number of
data as shown in Fig. 2, we have constructed a potential energy function for two
* ground state helium atoms interacting at -1 A. There are numerous theoretical and
experimental estimates for this potential function (see Foreman, Roland, and Coffin4)
all of which differ by more than 1.5 eV. Our preliminary measurements favor the
lower of all of these potentials.
Figure 3 shows a high-resolution scan of the low-energy two peaks for various4He- ion energies. The kinematic shifts clearly show that the two peaks which
result from forward and backward electron ejection in the center-of-mass system
corresponding to 11.4 meV autodetachment energy. The potential energy function for
He- (41") is approximately congruent with that of He; (a3Z,+) and shifted down
by approximately 0.223 ± 0.030 eV according to the calculations of Michels. 3 As a
result, vibrational autodetachment of excited He 2 is expected to follow the propensity
rule AP = 1 and AJ = 0. Therefore, except for the small anharmonicities
(which are too small to detect at the present level of resolution), single, narrow peaks
are expected for both ejected electron directions, as shown in Fig. 3. Figure 4 shows
the same spectra for the isotopic species 3He2. The autodetachment peak shifts up
to 40 meV in the center-of-mass frame. Again neglecting small anharmonicities, the
- difference between the autodetached electron energy for 3He2 and 4He- would begiven by 3/2(h,- - hv) - l/2(hp, - hP° ) where v- and v° represents
the frequencies for 3He ° and 4He, respectively. Using the measured difference of
37.4 - 11.4 = 26 meV and known values for hv, and hV4, we calculate a vibrational
energy of 0.22 eV for 3He- and 0.019 for 4He- which is close to that calculated
by Michels. As expected, the negative ion frequencies are very close to those of the
corresponding a3Z state. Assuming that these peaks represent vibrational
autodetachment from v = 1 of the ion to v = 0 of the neutral, we calculate an
electron affinity for 4He2 of 0.18 eV which is - 0.05 eV lower than the theoretical
* value of Michels who, in fact, computes the first vibrational level of He2 ( 4fl) to be
Monoenergetic Electron Source Using Autodetachment from Metastable Be
A simple apparatus has been constructed that should allow ultrahigh (<1 meV)
resolution electron spectroscopy studies. The method utilizes that fact that
Be'(ls 2s2p ) negative ions are metastable and eject monoenergetic electrons. In this
connection we have constructed a simple gas cell in order to study electron
transmission spectroscopy in the rare gases as a diagnostic test of the method. This is
a minor, but perhaps important, deviation from our studies of metastable negative ions.
This represents only a partial list of experiments we plan for the coming year.
Parallel to these efforts J. S. Thompson has been designing an apparatus which will
allow photodetchment studies of the negative ion beam discussed above. His studies
will be similar to our earlier studies5 of photodetachment of He- where total
detachment cross sections were determined. This work along with the autodetachment
studies will constitute his Ph.D. dissertation.
During the coming year we will continue our progress on studies of electron
collisions in a laser field. Dr. T. J. Kvale will devote part of his time to these
experiments. This should hasten progress in this area considerably. The basic
experiments which are underway have been described in the previous proposal and will
not be restated here. In addition to these experiments during the past year, there has
been considerable interest in the so-called "above threshold ionization" (AT)
spectroscopy or "continuum-continuum" scattering. We have added new information to
this exciting field (see Appendix I). Our studies were limited to alkali atoms using
laser power densities of -- 10I W/cm 2 . Recently, there has appeared studies on
xenon atoms by the Dutch group headed by M. Van der Weil and a French group
headed by G. Mainfray. These results are somewhat in discord with regards to the
ATI data. We intend to repeat these measurements and also the measure the angular
distributions for the ATI electrons.
r_. . . . . .~. . . . .
.[. ."..
10
References
1. See G. D. Alton, R. N. Compton, and D. J1. Pegg, Phys. Rev.A 28, 1405 (1983)
and other cited therein.
2. Y. K. Bae, M. I. Coggiola, and I. R. Peterson, Phys. Rev. Left. 52, 747 (1984).
3. H. H. Michels, Plays. Rev. Lett. 52, 1413 (1984).
4. P. B. Foreman, P. K. Roland, and K. P. Coffin, J. Chem. Plays. 61, 1658 (1974).
5. R. N. Compton, G. D. Alton, and D. I. Pegg, J. Phys. B 13, L651 (1980).
!=A. "
Table 1. Summary of experimental and theoretical values for theenergy levels of the Be-( s22s2p2)4P state
Electron affinity State energyE[Be('P) - Be-(4P)1 E[Be'(4P)]
(meV) (eV) Type Reference
240 ± 100 2.49a Theoretical I
> 122 2 .6 6 b Theoretical 2
285 Theoretical 3
217.7 57.1 2.56c Theoretical 4
195 90d 2.53 ± 0.09 Experimental Present work
aused - 14.6189 a.u. for the ground state of beryllium.
bused - 14.6684 a.u. for the ground state of berylliumas quoted from Ref. 5.
Cused - 14.667328 a.u. for the ground state energy of beryllium
as quoted from Ref. 5.
dused 2.7248 eV as the state energy of Be( ls'2s2p)3P° taken from Ref. 6.
Budget
A. Support for Personnel (Including Fringe Benefits and Overhead)
1985 1986
R. N. Compton(5%) $10,000 $12,000
G. D. Alton (5%) 10,000 12,000
Postdoctoral Fellow 41,000 41,000
B. Ancillary Cost
Technical Support 6,000 6,000
Travel 1,000 1,000
Channeltron Detector 1,000 1,000
Publication and Supplies 1,000 1,000
TOTAL 70,000 74,000
.........*. . . . . . .
* * *...
APPENDIX I
RESUME
NAME: Thomas Jay Kvale
TITLE: Postdoctoral Fellow in Physics
EDUCATION:
A.B. Gustavus Adolphus College (Cum Laude), 1975M.S. University of Missouri-Rolla, 1977Ph.D. University of Missouri-Rolla, 1984
ACADEMIC AND PROFESSIONAL EXPERIENCE:
1974-1975 Academic Assistant, Gustavus Adolphus College1975-1977 Graduate Teaching Assistant, University of Missouri-Rolla1977-1983 Graduate Research Assistant, University of Missouri-Rolla1984-Present Postdoctoral Research Fellowship with Oak Ridge Associated
Universities at the Oak Ridge National Laboratory
Ph.D. DISSERTATION TITLE:Angular Differential Cross Sections for the Excitation of I'S Heliumto the 2'S and 21P States by 25- to I00-keV Proton Impact
PROFESSIONAL AFFILIATIONS:
American Physical Society, Division of Atomic, Molecular, and Optical PhysicsSigma Xi
1. J. E. Aldag, J. L. Peacher, P. J. Martin, V. C Sutcliffe, J. George. E. Redd. T. J. Kvale,D. M. Blankenship, and J. T. Park, 'Angular Differential and Total Cross Sections for theExcitation of Atomic Hydrogen to Its n=2 State by Helium Ions,' Phys. Rev. A 23, 1062 (1981).
2. P. J. Martin, K. Arnett, D. M. Blankenship, T. J. Kvalc, J. L. Peacher, E. Redd, V. C.Sutcliffe, C. D. Lin, J. H. McGuire, and J. T. Park, 'Differential Cross Sections for ElectronCapture from Helium by 25- to 100-keV Incident Protons,' Phys. Rev. A 23, 2858 (1981).
3. P. J. Martin, D. M. Blankenship, T. J. Kvale, E. Redd, J. L. Peacher, and J. T. Park,'Electron Capture at Very Small Scattering Angles from Atomic Hydrogen by 25 - 125-keVProtons,* Phys. Rev. A 23, 3357 (1981).
4. J. L. Peacher, T. J. Kvale, E. Redd, P. J. Martin, D. M. Blankenship, E. Rifle, V. C. Sutcliffe,and J. T. Park, *Elastic Differential Cross Sections for Small-Angle Scattering of 25-, 50-, and100-keV Protons by Helium Atoms," Phys. Rev. A 26, 2476 (1982).
5. E. Rille, R. E. Olson, J. L. Peacher, D. M. Blankenship, T. J. Kvale, E. Redd, andJ. T. Park, "Isotope Effect in Electron-Capture Differential Cross Sections at IntermediateEnergies,' Phys. Rev. Lett. 49, 1819 (1982).
6. E. Rille, J. L. Peacher, T. J. Kvale, E. Redd, D. M. Blankenship, and J. T. Park, 'MomentumTransfer Scaling in Hydrogen-Isotope Collision Systems,* Phys. Rev. A 27, 3369 (1983).
7. E. Rilile, J. L. Peacher, D. G. Seely, T. J. Kvale, E. Redd, D. M. Blankenship, and J. T. Park,"Isotope Effect and Momentum-Transfer Scaling in the Elastic-Scattering Differential CrossSections for Hydrogen-Isotope Collision Systems," Phys. Rev. A 28, 3642 (1983).
8. J. T. Park, E. Redd, T. J. Kvale, and E. Rille, 'Reactive Scattering Cell for Atomic Hydrogen andDeuterium,* Rev. Sci. Instrum. 54, 1247 (1983).
9. J. L. Peacher, P. J. Martin, D. G. Sely, J. E. Aldag, T. J. Kvale, E. Redd, D. M. Blankenship,V. C. Sutcliffe, and J. T. Park, *Angular Differential and Total Cross Sections for the Excitationof Atomic Hydrogen to Its n-2 Level by 25-150 Hydrogen Molecular Ions,* Phys. Rev. A 30, 729(1984).
10. T. J. Kvale, D. G. Seely, D. M. Blankenship, E. Redd, T. J. Gay, M. Kimura, E. Rille, J. L.Peacher, and J. T. Park, 'Angular Differential Cross Sections for the Excitation of I'S Helium tothe 2'S and 2'P States by 25- to 100-keV Proton Impact,' Phys. Rev. A 32 (1985).
II. T. J. Kvale, G. D. Alton, R. N. Compton, D. J. Pegg, and J. S. Thompson, 'An ExperimentalDetermination of the Energy Level of Be-( s22s2p 2)lPj,' Phys. Rev. Let. 55, 484 (1985).
12. G. D. Alton, T. J. Kvale, R. N. Compton. D. J. Pegg, and J. S. Thompson, "The Production ofCa- Through Sequential Charge Exchange with Li Vapor,' Nucl. Instrum. Methods (submittedfor publication).
13. T. J. Kvale, R. N. Compton, G. D. Alton, J. S. Thompson, and D. J. Pegg, 'AutodetachmentSpectroscopy of Metastable He2," Phys. Rev. Lett. (to be submitted for publication).
- . *
Kvalc
NONREFEREED PAPERS
1. P. J. Martin, J. E. Aldag, T. J. Kvale, J. L. Peacher, E. Redd, V. C. Sutcliffe, and J. T. Park,*Differential Excitation Cross Section Measurements of Atomic Hydrogen to the n-2 State by 25to 150 keV Hydrogen Molecular Ion Impact," Bull. Am. Phys. Soc. 23, 1087 (1978).
2. J. E. Aldag, J. M. George, T. J. Kvale, P. J. Martin, J. L. Peacher, V. C.Sutcliffe, andJ. T. Park, "Excitation of Atomic Hydrogen to the n-2 State by 15 - 175 keV Helium IonImpact." Bull. Am. Phys. Soc. 23, 1087 (1978).
3. P. J. Martin, T. J. Kvale, J. L. Peacher, E. Redd, V. C. Sutcliffe, and J. T. Park. "DifferentialCross Sections for Electron Capture from Helium by 50 keV Protons," Bull. Am. Phys. Soc. 24,1187 (1979).
4. T. J. Kvale, J. L. Peacher, E. Redd, P. J. Martin, D. M. Blankenship, V. C. Sutcliffe, and J. T.Park, "Differential Elastic Scattering Cross Sections for 25, 50, and 100 keV Protons Incident onHelium," Bull. Am. Phys. Soc. 25, 1116 (1980). (Presented conference talk)
5. E. Redd, J. L. Poacher, T. J. Kvale, P. J. Martin, D. M. Blankenship, and J. T. Park,"Structure in the Differential Cross Section for Excitation of Atomic Hydrogen to the n= 2 Stateby N' with a Velocity of 0.25 a.u.," Bull. Am. Phys. Soc. 25, 1116 (1980).
6. P. J. Martin, T. J. Kvale, J. L. Peacher, E. Redd, D. M. Blankenship, K. Arnett, andJ. T. Park, "Differential Cross Sections for Electron Capture in Proton-Hydrogen AtomCollisions," Bull. Am. Phys. Soc. 25, 1135 (1980).
7. J. L. Peacher, D. M. Blankenship, T. 1. Kvale, E. Redd, P. J. Martin,and J. T. Park,"Differential Elastic Scattering Cross Sections for Sodium Ions Incident on Atomic Hydrogen forkeV Energies," Proceedings of XII International Conference on the Physics of Electronic andAtomic Collisions (Abstracts of Contributed Papers, Sheldon Datz, Ed.), Vol. 1, pp. 524-525(1981).
8. P. J. Martin, D. M. Blankenship, T. J. Kvale, E. Redd, J. L. Peacher, and J. T. Park,"Projectile Structure Effects in Electron Capture at Very Small Scattering Angles for(Is 22s22p")+ + H(ls) Collisions at 55.6 keV," Proceedings of XII International Conference on thePhysics of Electronic and Atomic Collisions (Abstracts of Contributed Papers, Sheldon Datz, Ed.),Vol. 2, pp. 659-660 (1981).
9. T. J. Kvale, E. Redd, D. M. Blankenship, P. J. Martin, J. L. Peacher, and J. T. Park, "AngularDifferential Cross Sections for Neon and Sodium Ion Impact Excitation of Atomic Hydrogen at anImpact Velocity of 1/3 a.u.," Proceedings of XII International Conference on the Physics ofElectronic and Atomic Collisions (Abstracts of Contributed Papers, Sheldon Datm Ed.), Vol. 2, pp.750-751 (1981). (Presented conference poster).
10. J. L. Park, E. Redd, T. J. Kvale, D. M. Blankenship, P. J. Martin, and J. L. Peacher, "AngularDifferential Cross Sections for Lithium Ion Impact Excitation of Atomic Hydrogen to the n--2State at a Relative Velocity of v - 1/2 a.u.," Bull. Am. Phys. Soc. 26, 1306 (1981).
IH. T. J. Kvale, J. L. Peacher, E. Redd, E. Rille, D. M. Blankenship. and J. T. Park, "MomentumTransfer Dependence in the Proton-Helium Collision System from 25- to 100-keV." Bull. Am.Phys. Soc. 28, 815 (1983). (Presented conference talk)
12. E. Rille, J. L. Peacher, D. M. Blankenship, T. J. Kvale, E. Redd, and J. T. Park. "DifferentialCross Sections of Hydrogen Isotopes at Small Scattering Angles," Bull Am Phvs. Soc. 28, 797(1983).
13. J. L. Peacher, D. G. Seelk. T. J. Kvale. E. Redd, D. M Blankenship. and J. T. Park."Excitation of Hydrogen to ItN n=2 Level by H 2 in the Glauber Approximation," Bull. Am.Phys. Soc. 28, 803 (1983).
14. E. Rille, J. L. Peacher, D. M. Blankenship, T. J. Kvale, E. Redd, and J. T. Park, "Hydrogen-Isotope Electron Capture Differential Cross Sections at Small Scattering Angles," Bull. Am. Phys.Soc. 28, 940 (1983).
15. J. T. Park, D. M. Blankenship, T. J. Kvale, J. L. Peacher, E. Redd, and E. Rille, "ElasticDifferential Cross Sections for Proton Scattering by Atomic Hydrogen," Proceedings of X1IInternational Conference on the Physics of Electronic and Atomic Collisions (Abstracts ofContributed Papers, J. Eichler, W. Fritsch, 1. V. Hertel, N. Stolterfoht, and U. Wille,Eds.), p. 350 (1983).
16. E. Rille, J. L. Peacher, T. J. Kvale, E. Redd, D. M. Blankenship, D. G. Scely, R. E. Olson, andJ. T. Park, "Small Angle Scattering in Hydrogen-Isotope Collision Systems at IntermediateEnergies," Proceedings of Symposium on Atomic and Surface Physics (SASP '84).
17. T. J. Kvale, D. G. Seely, D. M. Blankenship, E. Redd, T. J. Gay, E. Rille, J. L. Peacher, andJ. T. Park, "Angular Differential Cross Sections for the Excitation of I'S Helium to the 2Sand 2P States by 25- to 100-keV Proton Impact," Bull. Am. Phys. Soc. 29, 777 (1984).(Presented conference talk.)
18. T. J. Kvale, D. G. Scely, D. M. Blankenship, E. Redd, T. J. Gay, E. Rille, J. L. Peacher, and J.T. Park, "The Excitation of Helium to the n=3 Composite Level by Proton Impact," Bull. Am.Phys. Soc. 29, 810 (1984). (Presented conference talk.)
19. E. Redd, D. M. Blankenship, T. J. Kvale, T. J. Gay, J. L. Peacher, and J. T. Park, "CoreExcitation in Mg + Ion Impact of He,' Bull. Am. Phys. Soc. 29, 812 (1984).
20. T. J. Gay, D. M. Blankenship, D. G. Seely, T. J. Kvale, J. T. Park, J. L. Peacher, and E. Redd,"Impact Excitation of He by He' and H+: Alignment of the He 2'P State," Bull. Am. Phys. Soc.29, 817 (1984).
21. G. D. Alton, T. J. Kvale, and D. J. Pegg, "Production of Ca- Through Double Charge Exchangewith Li Vapor," ORNL Physics Division Report ORNL-6120, p. 11 (1985).
22. G. D. Alton and T. J. Kvale, "The Negative Ion Source Test Facility as a Negative Ion AtomicPhysics Research Facility," ORNL Physics Division Report ORNL-6120, p. 138 (1985).
23. G. D. Alton, T. J. Kvale, R. N. Compton, D. J. Pegg, and J. S. Thompson, "Progress TowardMeasurement of the Properties of Metastably Bound Negative Ions," ORNL Physics DivisionReport ORNL-6120, p. 139 (1985).
24. T. J. Kvale, G. D. Alton, R. N. Compton, D. J. Pegg, and J. S. Thompson, "An ExperimentalInvestigation of the Metastable States in Beryllium Anions," Proceedings of XIV InternationalConference on the Physics of Electronic and Atomic Collisioas, (Abstracts of Contributed Papers.)M. J. Coggiola, D. L. Huestis, and R. P. Saxon, Eds.), p. 415 (1985). (Presented conferenceposter.)
Experimental Determination of the Energy Level of Be- ( 1s2s2p1)4P
T. J. Kvale, G. D. Alton. and R. N. Comptona)OaA Ridge National Laboraton. Oak Ridge, Tennessee 37831
and
D. J. Peggi b ' and J. S. ThompsonThew Lnver.str of T(nnes.se. Knovvihle 7'nnesssee 3 7996
(Received II March 1985)
We report the first experimental measurements for the energy level of a metastable state of Be-The ions were produced in sequential charge-exchange collisions between 50- to 60-keV Be* ionsand lithium vapor The center-of-mass energy of" autodetaching electrons was found to be2 53 ±0 09 eV This result is in good agreement with previously calculated values for theBe - ( I s 2Q 112 )4P-state energy
PA(S numbers 32.80.1)z, 35 1O1n
In this paper, we report the observation of a peak in cur at either 263.8 nm (Ref. 9). 267.1 nm (Ref. 12),the Be autodetachment electron-energy spectrum or 265.4 nm (Ref. 10). Radiation from this transitionwhich is a signature of the decay of a metastable beryl- was searched for by Andersen 7 without success. Alium negative-ion state. These measurements third metastable state, Be-(ls2s2p3 )bSo, has alsorepresent the first time that the energy level of a been predicted.9'I 2 This state, however, lies energeti-long-lived metastable state has been experimentally cally outside the present experimental range [ > 100determined for negative ions of the group IlI eV from Be(ls22s2 )tS]. Other Be- ion states have(alkaline-earth) elements. In fact, limited experimen- also been theoretically studied. For instance, thetal information is available on the structure of any Be-(I s22s2 2p) 2p configuration is predictedt8 to be ametastably bound atomic negative ion. Notable excep- shape resonance; however, the lifetime of this state istions to this include lie- which is .i classic example of too short to be studied with the present apparatus.an spin-aligned metastable negative ion (see, e.g., A(- Theoretical calculations of the structure of negativeton, Compton, and Pegg' ), resonance studies from ions are particularly difficult since the electron affini-electron-atom scattering experiments (see, e.g., Bur- ties are typically of the same magnitude as the differ-row, Michejda, and Comer:). and Li-, in which pho- ence in correlation energies between the atom and ion.ton emission was observed between high-lying meta- Even so, most of the present information concerningstable states of the negativc ion (see, e.g., Brooks metastable negative ions, other than He-, has been -.
et a/. 3). The fundamental nature of Be - makes it of provided by theoretical studies of open-shell excited-considerable experimental and theoretical interest. As state negative-ion configurations. The first theoretical
, early as 1966. the ion was reported4 to be present in investigations of the structure and binding energiesmass spectra of ions emitted from direct-extraction (electron affinities) for the group IA elements Li-,negative-ion sources. Since the first observation of Na- and K- and the group ilA elements Be andBe-. experimental values for production efficiencie-' Mg- were made by Weiss8 who employed variational-and autodetachment lifetimes' have been reported In superposition techniques. The Be- ion was predictedrecent measurements on the autodetaching decay of to be bound relative to the neutral atomicBe Bae and Peterson' have shown that the ion has at (ls 2 2s2p)3P * state by 240 meV with the most likelyleast two distinct lifetime components of _ 10- 4 and configuration postulated to be Be-(I s22s2p 2)4P. This
-10- 5 s. configuration is metastable against autodetachmentMost theoretical studies indicate that Be is meta- by spin-forbidden transitions-analogous to the
stable,8-l' although early theoretical calculations14- 6 (Is2s2p) 4P * state of He-.suggest that Be II s22s,3.s )2S is bound with respect to Configuration-interaction calculations have beenthe ground state. Is2s2)IS, of beryllium. Recent employed by Bunge et al.9 in a search for possibletheoretical calculations do not predict the existence of bound excited negative-ion state configurations for thea stable Be state but predict the existence of two elements hydrogen through calcium. The results ofstates, both metastable against autodetachment, which these investigations indicate the existence of two me-lie below the first ionization threshold of neutral tastable states of Be--the Be-(I s22s2p2 )4 which isherllium-Be- (I s2.s2p 2)4 P which is bound with bound relative to Be(ls 22s2p)3P by 285 meV, and" respect to Be(I22Sp )lP ° , and Be-(1.' 2l2p3)4S° Be-(ls2 2p3)4 S° which is bound relative to the
which is hound with respect to Be( I I s22p 2)3Patomic state by 262 meV.diatively allowed 'S' - 'P transition is predicted to oc- The fine and hyperfine energy separations, as well as
electron affinities of the lowest two bound states of Electrons ejected in the forward direction followingBe-, have been calculated by Beck and Nicolaidesi ° autodetachment or collisional stripping were energythrough the use of a nonrelativistic many-body calcula- analyzed by the electron spectrometer which wastion. These studies indicate that Be - ( I s22s2p2 ) 4P is operated in the fixed-pass-energy mode. The col-bound by 218 meV relative to Be( Is 2 2s2p) 3P ° , while lisionally detached peak served as the absolute refer-Be- (ls 2 2p 3 )4S° is bound by 244 meV relative to ence for the determination of the energ of the auto-Be I.s22p 2 )3P. The lifetimes of the Be -( 41j) levels detached electron peak in each spectrum. The ionwere also predicted'3 in a recent calculation. beam energies were chosen sufficientl% high (50, 55,
The experimental techniques utilied for determin- and 60 keV) so that the collisionallv detached electronIng the energy levels of metastable Be- arc very silni- peak could he resolved from the lo-energy electronlar to those used for the measurement' of the energy background peak centered at iero laborator, energy.level of fie l 2s 2p) 4P*. For the present investiga- Autodetached electrons ejected in the backNard direc-tion. a Be' in beam "a , post-acclerated to the tion could not he resolved from this lo% -electron-,hosen kinetic energy, nonCntUn anal ied, and energy "noise- peak. [he electron signal iron) theh)co,,a through I lithium vapor cc!] situ,ited I m ,.kj% anilyzer Nas input to a ('A IA('-based multichannel-trorm the electron spectrometer as shown in Fig. I analver data-acquistion system. Each channel of the-his resulted in a time delay of - I ps between the multichannel analy/er corresponded to a unique elec-Be- production and detectioi of electrons ejected tron energy, and the typical voltage increment usedfrom the decay oI Be - Thus ontlv ions produced in during these measurements was 24 24 meV.states metastable against both autodetachment and ra- The center-of-mass (or state) energy E, n, of auto-diation surxive long enough to be experimentally stud- detached electrons ejected in the forward directionted. The positive, neutral, and negative components from a moving ion beam can be determined relative toof' particles emergent from the lithium vapor cell were the electrons which move at the projectile velocityseparated upon entrance into the experimental (collisionally detached) by use of the small-anglechamber by an electrostatic deflector. Ions passing kinematic expression (see Ref. I). The resulting rela-through the device were deflected by + 10' into a tion for E. is given byI 5-cm-long gas cell. The purpose of the gas cell Aasti) produce a high-pressure region f'or collisional strip- U m = - x -]. (I-
ping of electrons from the Be- ion beam. .\ smalliperature located in f'ront of the gas cell served to col- whre f-E, is the laborator% energy of the atodetat hinglImate the Be- beam prior to its passage through the peak and I1K is the laborator, energy (if the c)lli,<onal
gas cell .ind through the 16(0' spherical-sector detachment '-'ak 1i is determined from the kinetic
Cl tron-enerey a Thc on beam. after energ., of the , beam B. use of th!, procedure tostraight-line transit through the analiyer, Aa,; moni- determine F:, ' .ih i[ based (il to peaks corre-tored in a hielded Faraday cup located it the rear of ,ponding to ph .,!,.:l c\cnt, occurring in the same
the spectrometer. Mutually perpendicular sets of speLtrum, errors duc '- contact ind surface potentialsllelmholt7 coils were used to nullift' the Earth', and are minimized and the "eed for precise kno\ledge (ifstray magnetic fields in the 'vicinity of the electron the spectrometer consta;;t is avoided.spectrometer. Electron energy spect,;i were acquired for Be- ion
COLLISIONALDETACHMENT CELL - CHANNEL ELECTRON MULTIPLLR
NEGTIV ELECTROSTATIC SPECTROMETERNEGATIVE , \ \
ION BEAM-- , , /-ELECTRON SEAM
DEFINING APERTURE-- F Y UP
ION EAM -'SECONDARY
J._. >FARADAY ELECTRON-- CUPS SUPPRESSOR
RECIRCULATING LiCHARGE EXCHANGECELL/ /
LSECONDARY ELECTRONPOSITIVE SUPPRESSORION BEAM
FIG. 1 The experimental arrangement used to measure the energy level of the metastable Be- () state
VOLtUME 55, NUMBER 5 PHYSICAL REVIEW LETTERS 24 Jt I N 1X5
introduced into the gas cell, while the autodetachmentpeak was present in all of spectra. The addition of gas
Ao,,detached ite -tron, in the cell does not change the shape or the energy po-tie. t(,,s s:' sition of the autodetachment peak. This peak. ,hichE, :.,:i v% occurs at the laboratory energy of 11.71 eV for 55-keV
Be- ions, is attributable to autodetaching electronsfrom the decay of the Be- (4p) state. This signal (typ-ically several kilohertz) was well resolved from theelectron background (typically < 100 tHz). while thesignal-to-noise (SIN) ratio of the collisional-
S,,detachment peak was typically less than that of the au-t,. ,, 1. todetachment peak. The lower Si.4 Nof the collisional-- . detachment peak was due primarilh to the highe; elec-
i o ,ri XrIron background at lower electron energies. Thepresent measurements gave an average value for thecenter-of-mass energy of this state measured withrespect to the ground state of neutral beryllium of 2.53
S "eV with a standard deviation of ±0.,9 eV. TheFWHM in the laboratory frame of the peak shown inFig. 2 is - 0.2 eV. which corresponds to a FWHM of
\ ,, rg,'t (,as - 0.04 eV in the center-of-mass frame. From theseresults, we estimate an electron affinity of 195 ±90
10.0 15.0 meV using the value from Bashkin and Stoner 9 forFit(L itrt 1'Ir,.v F 4 (,,V) the energy of Be( Is 22s2p)-P ° . A search for the elec-
I I( 2 Electron encrg. spectra trom 55-keV Be- ions tron autodetachment peak from Be- ( Is 22p 3 )4S was% ith and , ithout a thin argon target made without success. This is not unexpected since
this state is permitted to decay radiatively to thelower-lying Be- ( I s22s2p2 )4 P state.
cnerg es ot 50. 5. and 6) kc:V Figure 2 displays a Table I compares the present results with previoushigh-rcsolut,,n autodcta:hment spectrum taken theoretical calculations. The electron affinit is de-without target gai, iod i ,pcctrum taken with - 2 fined as the energy difference between Be( 3 P) andrnilorr argon in the gas ,.I I to enhance the collisional- Be- (4 P), whereas the energy separation betweendcta,.hrncnt peak. 1 he ,pec Ira were taken sequentlall% Be(tS) and Be- (4p) is the state energy of Be -( 4 p)and at an ion -nerg) off §5 keV. The very low,- In cases where the total energy of the ground statecncrgy-clectron background peak is attributable to Be(0S) was not reported. the latest reported ',alue b.,o.-cncrg electrons generated by ion impact with Bunge20 was used in order to arrive at the Be- (4 p IA.perturcs. etc The collisional-detachment peak was state energy. In both cases, the resulting energy is aonly detectable whenever a target gas (e.g., argon) was small difference between rather large numbers, and er-
TABLF I Summary of experimental and theoretical values for the energy levels of theBe ( 2 s 2p')'P state.
Electron affinity State energyE(Be( 'P) )-E(Be- (P)) E(Be-(4P))
'Used - 14 6189 a.u for the ground state of beryllium
'Used - 146684 a.u for the ground state of herlium. as quoted from Bunge iRef 20)'Used - 14.667328 a u for the ground-state energy of hervlium, as quoted from Ref 20dUsed 2 27248 eV as the state energy of Be(1 ,2% 2pi1P, taken from Ref 19
F 486
VOIA.%i: 55, N lilt.R5 PHYSICAL REVIEW LETTERS 29 JLutN 1985
rors. uncertainties, or omissions are extremely impor- (41Also with the t)artment of Chemistry. Lnlxersit% oftant. For example, calculations (see, e.g.. Ref'. 8) Tennessee. Knoxville, lenn. 37996.which do not take electron correlation effects into ac- '6iAlso with Oak Ridge National Laboratory. Oak Ridge.count predict be-(P to be unbound (electron affini- Ten. 37.loRN.Cmtnad1)J.PghsRety = - 68 meV). The present electron-energy mea- GI.Ato.RN.(mtnadI)J.eg.Py.Rxsurements covered the whole region below the firs DA Burow 140 A(icda9nd8Conr.3)Py. 9
ionization threshold of neutral beryllium, and only one 3225 ( 1976).autodetaching peak was observed in the spectrum. 3R. L. Brooks. J. F. Ilardis. 11. Gi. Berry. I . J. Curtis. K. THowever, this does not conclusively rule out the ex- Cheng. and W. Rax. Phs Rex. Lett. 45. 1318 (1980). .istence of other metastably hound states (or Be since 4K. Bethge. L-. Ileinicke. and II. Baumann, Phys. [Lett. 23. -our apparatus is insensitive to ions having lifetimes 542 (1966).much less than I jzs or much greater than at few tens of' 5[ Ileinicke. K Bethge. and 11 Baumann. Nucl. Instrumn.microseconds. This includes, of course. ion states that Method,, 58. 125 (1968).are not metastable against radiative decay. 6J. lleinerneier and P. livelplund. Nuci. Instrum. Methods
In conclusion, the present results support the 148. 65 (1978). J. lleinemecier and P). llCeirdUnd. Nuci. In-strum. Methods 148. 425 (1978).theoretical predictions that there is only one state of YK.BeadJRPtronPhsRxA3.24
the negative beryllium ion lying below the (184. K a n .RPtroP~ eA3,24Be( Is 2s2p ))Othreshold which is metastable against 8A. W. Weiss. Phvs Re%. 166, 70 (1968Vboth autodetachment and radiative decay. The agree- 9C. F. Bunge. M. (Galan. R JKiiregui. and A' V Hunge.ment between the theoretically predicted electron- Nucl. Instrum. Methods 202. 299 (1983),affinity and state-energy values and our experimentally 10 1). R Beck and C. A Nicolaides. nt. J Quantumn (hemndetermined values permits us to identify the observed Quantum Chem S\ mpl. 18. 467 (1984)autodetaching state as Be-(] .x22s2p2 )4I. Additional 11C A Nicolaides. Nt Komininos. and 1) R Beck. Phxssupport for this identification comes from lifetime pre- Re\,. A. 24. 1103 (1981)dictions' and m~easurements~ which indicate that 2) ek .\NcIids idC srn~li.lhBe -
41 Iwill have a lifetime component matched to RxA2.022(91
141C Aspromallis. ( A~ Nicolaides. and 1) R Beck (to hethe time window of'our apparatus. . publshed).
This research was sponsored by' the [Division of Basic 4 ole.I hmP~ 0 21I1 6 )
Lnerg Sciences. U. S. Deparimemn of Energy under 15H. S. W. Massey, Nf'gatoIv'on.s (Cambridge L nix. Press,Contract No. De-A('05-840R2 1400 with Martin Cambridge. 1976).Marietta Energy Systems. Inc.. the U. S. D~epartment 16B. L. Moiseiwitsch, in .- tomt, Pn und -I pp/ftca/10%.of' Navy., Office of' Naval Research. Task 393-071. and edited by P. CG. Burke and B. L. Moiseiwitsch (North-the U. S. Department of' Energy through the Unixersi- Holland, Amsterdaim, 1976).x' of' Tennessee under Contract No. DE--AS05- t 1T Andersen. private communication.
83ER 13097. One of us (1.J.K.) is at postdoctoral Eel- 1811. A Kurt/ and Y. Ohm., Phivs. Re\ .\ A19. 43 (1979)low in the U. S. Department of Energy Postgraduate 19.4 romi Energyi Levels and G;rowna~ I),aira1%s IP Hidrogeo(
Research Tratning Program administered by Oak I-Phosphorus AT,. edited by Stanley Bashki -, and John 0.Ridg AsocitedU niveri tes.Stoner (North- Holland, Amsterdam. 1975).Rtde Asocate Unveritis. 20C. F Buinge. Phys. Rev. A 14. 1965 (1976).
487
" - . -.
4 PHYSICAL REVIEW A VOLUME 31, NUMBER 4 APRIL 1985
Electron-detachment spectroscopy of 20-100-keY H- projectiles interactingwith thin Ar targets
G. D. Alton and R. N. ComptonOak Ridge National Laboratory. Oak Ridge, Tennessee 37830
D. J. PeggDepartment of Physics. Uniiersity of Tennessee, Knoxville, Tennessee 37916
and Oak Ridge National Laboratory. Oak Ridge. Tennessee 37830
(Received 30 July 1984)
Energy and transformed-velocity spectra of electrons emitted in the forward direction have beenexperimentally determined for 20- 120-keV H projectiles traversing thin Ar targets. The spectrafiom H are surprisingly cusplike in shape even though single-electron loss is known to dominate.In contrast to the findings of Menendez and Duncan. who performed similar measurements athigher projectile %elocities, target-induced structures in the electron loss to the continuum spectraare not pronounced in this %elocity regime.
1. INTRODUCTION projectile velocities at the peak of the distribution. Insuch ion-atom collisions, electrons are known to originate
Considerable efforts have been expended by several from capture to low-lying, projectile-centered states forresearch groups toward elucidation of the complex physi- bare or nearly bare energetic projectiles (ECC) and fromcal processes of electron capture to (ECC) and loss from loss to low-lying continuum states whenever loosely bound*ELCI continuum projectile states, both of which occur as electrons are present on the projectile (ELC). Consequent-a consequence of their traversal through thin gaseous and ly, for partially stripped multiply charged ions, both pro-solid targets. These physical phenomena, though distinct- cesses may be present-making interpretation of the re-ly different, both exhibit themselves as sharp, cusplike suiting spectra difficult or impossible.peaks in the ejected-electron spectra whenever the electron The results of early theoretical approximations of thevelocity v, matches the velocity of the ion in speed and proton-induced target-electron-ejection phenomenon baseddirection. The spectral positions and similarities in shape on the plane-wave Born 6 and binary-encounter models7
of ECC and ELC processes make experimental distin- were found to be in good agreement with differentialguishability extremely difficult or impossible in cases cross-section measurements at intermediate angles. How-where both processes may be present. For these reasons ever, at very small and large angles, significant discrepan-brief comments regarding previous studies of ECC are cies were observed. Thus, neither model could explain thealso included. The reader is referred to the articles by Sel- enhanced doubly differential cross sections observed atlin' and Breinig et al.2 for more comprehensive reviews small angles by Rudd et al.3 The breakdown of theseof previous investigations of these phenomena. models for forward-ejected (ECC) electrons is attributable
Evidence of the ECC phenomenon was first observed to the exclusion of long-range attractive projectile-electronby Rudd et al. 3 in the form of enhanced doubly differen- interactions-a critically important requirement for ex-tial cross sections for electron ejection in the forward planation of this effect.6 That is, the interaction must notdirection from atoms subjected to proton bombardment, only transfer momentum to the ejected electrons but mustSubsequent measurements by Crooks and Rudd4 of the include postcollision long-range effects due to the attrac-electron energy distributions resulting from the bombard- tive forces between the ejected electron and the positivelyment of He targets, also with H" projectiles, revealed charged projectile. Several theoretical efforts have beensharp, cusplike spectra with the electron velocity at the made which take into account these final-state interac-peak of the distribution very close to that of the incident tions including the works of Salin, Macek," Dettmann,projectile. Similarly shaped peaks were also observed by Harrison and Lucas, 3 and Garibotti and Miraglia.t 1Harrison and Lucas5 in the momentum distributions of The physical basis of ECC was successfully explainedelectrons emitted from thin carbon foils in the direction of by Maceki through the use of a multiple-amplitude first-the incident proton beam. order approximation to the Faddeev integral equations
Since these early discoveries of the ECC phenomenon, with the final state for the ejected electron represented bymany investigations have been made of the energy and an attractive target-centered Coulomb wave function.momentum distributions of electrons resulting from in- The theory predicts an approximately symmetric electronteractions between a variety of fully stripped light and distribution in space at ve -v, with a yield proportional tomultiply charged ions interacting with thin gaseous and Z', where Z is the charge on the projectile. Othersolid targets: all have confirmed the cusplike nature of the theoretical efforts to describe the ECC phenomenon inresulting spectra and the equivalence of the electron and terms of single-amplitude first and second Born approxi-
31 2129 c 1985 The American Physical Society
2130 G. D. ALTON, R. N. COMPTON, AND D. J. PEGG 31
mallions have been made, including the first Born, exhibits itself in thle Form of an electron distribution. simi-*Coulomb-distorted wkave description of proton ionization lar in shape to that of' ECC with the peak centered at
of' He by Salin' which predicts an asymmetric peak in the v, v,. The origin of the peak canl be readily isualized* selocity spectrum of ejected electrons at low% projectile ke- tin terms of an energy or velocity distribution inmparted to
locities. Dettmann, Harrison, and Lucas' have applied the detached electrons superposed ahout the centroid of'rithe second-order Born approximation for comparison of the moving projectile. The underlying theoretical basis of'
theoretical shpe adyesof CC spectra with those ELC was made initially by Drepper and Briggs1 s . Io ap--xeieal bee o ,H ,aH - ions pass- plied the first-order Born approximation wvith projectile-
Ing through thn carbon foils. The theory, predicts a Z' centered attractie Coulomb wkave functions for descrip-yield dependence and svnmetric cusp-shaped spectrum tion of' projectile ionization fo elcrn mos in inthcentered at v,-v, at high projectile velocities with the beam direction. Briggs and Drepper. tin a lter article,full] wvidth i at half' maximutm fFXH M F increasing linear- used t hc fi rst Born approximation to calculate the crossI\ lxssithl projectile %elocit\ v, according to F - i 01 ,. section. differential in angle and s elocity . for projectile-\\where 0:, is the hal f-angle about thle direct ion of'projectile: elect ron loss. Thle theory predicts at cusplike \eloci iv dis-Miottin. Ho%%eser. first -order si ugle-amplit ode theories t ribution for electrons mos me in t hc forsvard direct ion55ic prdc mmlllctrie spectrat for electrons emitted it w\hile electrons ejected in the backward direction arele forss ard direction at hiigh projectile eclocit tes (to not found to be adeq uatel\ characreri zed b\ ant elastic scatter-
~Igree %%lit thle asx inieitrnc character experimentall\ ob- ig moidel. The theory predicts cross sections for this pro-Ners ed b" Vanle (,I i? or lull, stripped carbon and ifir cess w~hich approc constant salues in the limit of'oss ,cu ions, interact ingls %kii .\r and h\ Rodbro and .\n - V .'V,, \%where v, is tile projectile seloci t and v,, that ofderseti for prot ns itract in t ie l ti lie. Ne. Ar. and H., the orbital velocity of the detached elect ron. This predic-
tarot I e as ii metnc h ra i N s c 11CC hi c01- tlion is .supported by the observed t rends, toward constant* s elcII\ sIpectra airc predlicted s\ hcs cr hiL'her-order cross sections da /dfl with increasing projectile %elocities
I[Cterms itL HIlIL-Id III tII i1,n es\parisio Sliakeshaft and made by Strong and Lucas.'~- Theoretical deriations of''sprich S tihseqUII ucboItlat Ot b% ( iarihot t and the doubly, differential ELC cross sect ioiis iii the firstNiiraelia 1111i10 11 l 1e I s let iCIII dens eLd frontit a rul- Born plane-wave approximation have also been made by
tiles~terti epls t the I iiatri\ also eild an Blriggs and Dax 1 and Da\. Tile dependence (if the crossIs ItIiet L' I I I k . Ic. t iii distribuition for tlt't> section onl smnall angles 0,, with respect to the direction of'-
trmu' cievcd III the t (t 11Co 11011e tille beamn is derived. The FWHM of' thle Cusp for this ap-A recent p11ihl"1arW 1, [itcejioc ,I a suniniari/es ile prowination satries according to F- 0,,it'1 I -13
reslis den e1%d 1111 .11 oC sTeisII' kc prinien~til research I. _32 where f3 is the arlisotropy parameter v ic.iprogram desIed to studieN o4 1 (( tind I I C neair v, v, generial. Ii afIncerion of v,. All of' the theoretical estima-itiduICed Kh\ l-eiet l ii It p1 chiar'd ions t r3\s rsitie tlions, of' FLC cross sect ions are \Aid at projectile veloci--11t1i1 1_,iseots targcis tiitd tie( atfloentis procese ascit ties V - V..(*k %Wssih higicte roeJLCIdCs pass"ing through anior- 'Thle Pres iousl discussed in'estigations have iris lsed
- phousi. pois r\ xia.111 il. Aiiid NI IILeL-cI rSst a IiC tic !t ei. I hie the use if' posit i c charged projectiles iii measurements, of'
perCl .as \\('11. IJUinCl ist rat her c niiprehiensiw rLnesic\% of' FCC arid FLC tin laboratory obsers ed energy or ecloci t vot her repori ed ss onk desI~ ted iF CC. FI , C. and aliogouls distributions. rThi series of insest i ations by Menlendelprijectile solid plicilonierli at 1high projectile eloctines. and Dllcarr -- hias utilized H projectiles tin meastir-[lie: stwdies sscre imade \%il iti asarietv oif' protectiles anti Iiig Fl.C differential cross sections, iii energy arid angle re---
proectalechage tats o e ancriresrane o t)7X.5 sulting fromi initeractionis ss ith thin rare-gas targets, and* Me'sL u ssithI and( ss it hoUt cuincideince requirements be- projectile etiergies of' 0.5 and I MeVs. These measure-
* t sscii itle enierget charge state aiid fors\ ard scattered inetits are of' particular relevance to tile present work.electro ns. he iii cstigal titis show i5.1le FCC spectra in P~erhaps t lit most iteresti rig aspect of' tile svork of
ehs\locity regime to be stronigl\ skewed toisar(t losser Meiientei and Duncan sas the discoser\ kif targt
- sclocities V, svile ELC in the -sanme regime appears to be dependent structutres in thle ELC spectra near 0 - 0 froint-approxiriiatel% ssnimnetric \k ith FWHM\ independent of' lie arid Ar targets. Th pectra consisted frsodsic*v, . III conit radistinction, thle elect roir spcctra acconipatix - peaks--a doininant peak centered at v, _ v, and a smallet
ing ion peniet rationr t hrough solid targets, in t his eriergx peak tdisplaced toward loss er s elocities, v, by 30 arid14range are found to be slight l\ skeved tossard higher dlee- 135eV, resptct is el\ for Hie arid Ar- taretis.2 The
ron s elticities while exhibiting s elocitI -in~dependent obsers ed etucrges difference hcetseni thle respective groups,sitlts-ati assuniptioii which is, at variance %%lttt rheoren Is found to be Intdependenit iif' It selocit\. The tfiscereCA predictions. Projectile atomnic-number dependeciie ats- of' utispeceret spectral featuores associated \%lit ELC prii-soiciated with the FCC differetial cross section in s eloci - cesses has, further stimulated expterimnital efforts, antI-.-ti ((/dc for at giveii selocity v,, skcrc found to scale ac'- gis eni iipetus to a iiumber of' theoretical papers, skith thecoirtdinig tol Z' antd thus subl-staintialls tcIesate front the objectis es of' forritilatirig a phxsical batsis for thle respee-Z / theoretical predictions of' \acek" antI Dettiniarii liar- rise obsers anions. Theoretical ssiirks include those oif
* risorli. ntI [ticas.' Fran, Wrighlt . antI (itiorti, Da\."' Naleck i and \fa-Elect ron loss to tilie Corntinurtm (1C.0 occturs tdurinrg cek./L and recent! h\ b Ponice antI Baragiola."' Thle
project ile-target intecractiont, whberieer looselN bound CICC- doubled-peaketl st ructutre is shovn tin rue first Born ap-frons are present onl the projectile. The ionization process proximation to be a ctonsequecite of' tiestructive iriterfer-
* 31 ELECTRON- DETACHMENT SPECTROSCOPY OF 20-100-keV ... 2131
encc betwecen I - 0 and higher I , 0 partial w\aNes or clec- those reported previously.- Negative ions were generatedIron excitation of the target. Such structure was also in a conventional duoplasmatron charge-exchange sourcepredicted by\ Garibotti and Miraglia in terms of an anti- through sequential charge-changing collisions with the ex-capture phenomenon for negative ions. More recently. change vapor of gaseous CH 4 . The desired final energy of'Burgd~rfer" has performed similar calculations which a particular negatixe-ion beam was achieved by adjusting"s.,ho\ a cusp in, ersion near v, _-v, for ELC processes in in combination or separately, the potentials for extractionan attractisc Coulomb field-the shape of which is deter- of the positive-ion beam and post acceleration of the
r. mined not old b\ the final-state interaction but by negative-ion beam. These potentials were accurately mea-d,,namical effects and b\ sym metry of the initial state as sured with a precision electrostatic voltmeter.wsell. The beams \ere momentum analyzed and focused into
A,, a continuation of the preiously reported studies of a biased and shielded Faraday cup located in the experi-the ,pectro,,co)pN of the metastablN bound states of le-.- mental chamber (see Fig. 1). A small hole located in the\%c ha\c measured ELC spectra near 0- 0 for HI projec- base of the cup served as a collimating aperture upicaltile, interacting with thin Ar targets at projectile energies diameter 0.75 mm) for the beam prior to entering abctssccn 2) and 120 kcV-a region of velocity in which 1.5-cm-long gas cell, terminated by 2 mm entrance andELC has not been previousl\ investigated. The principal exit apertures. The cell was used to produce a high-objecti\es of this investigation were to st udy the shape pressure region for collisionally detaching electrons fromand clocit, dependence of electron loss spectra from H the respectix e ion beam. Target pressures were monitoredprofctIes interacting with gaseous targects and to search .kith a conventional capacitance manometer. Care wasfor ,t rctures such as discovered by Menendez and Dun- taken during all measurements of electron energy spectracMn or collisionall\ induced autodetaching leels of to ensure operation within the cell pressure range forfhe,,c ions. Therefore, emphasis was not placed on mea- single-collision conditions. The range for a particular ion'Uremcnt of absolute doubl\ differential cross sections. species, detachment gas, and ion energy, was established
by arying the cell pressure and noting changes in height
II. EXPERIMENTAL APPARATUS of the detachment peaks. All pressures were then main-AND MEASUREMENT TECHNIQUES tained within the linear pressure range established for a
particular energy and detachment target.The detail, otf the experimental apparatus and tech- The ion beam and detached electrons, moving collinear-
niques utiliued in determining the energy spectra from ly with or ejected in the forward direction by autodetach-colli,,onall.> and autodctached 11 ions are %erv similar to ment in the 5-cm drift space following the cell. enter
ll(i. SCtle ltCI " drsIi Ilg A eile Clctron spcctro,,cop apparatus.
2132 G. D. ALTON, R. N. COMPTON, AND D. J. PEGG 31I-.
along the central ray into a high-resolution, double- from the assumed spectrometer resolution function andfocusing, 1800, spherical-sector, electrostatic analyzer with for velocity space transformation. The background con-a mean radius of 4 cm. The ion beam passes through the sisted of a monotonically decreasing electron signal ofspectrometer, exits through a high-transmission-gridded low-energy electrons centered about ve =0 in the laborato-aperture located at the rear of the analyzer and is moni- ry frame produced by the ion beam striking aperturestored in a shielded and biased Faraday cup. The beam situated at the entrance to the analyzer and a smaller sig-collimation used in the experiments permitted angular ac- nal which was attributable to electrons originating fromceptances of 0: 1.50 with respect to the central ray of the the target due to direct ionization. The data shown inentrance aperture of the analyzer. Fig. 2 illustrate the nature of the background spectra and
During all energy-distribution measurements, the elec- are typical of those observed in these experiments. Atron spectrometer was operated in a fixed pass-energy monotonically decreasing function, similar to that shownmode thus requiring the acceleration of electrons moving in Fig. 2, was used to separate the ELC and low-energyat energies less than and deceleration of those moving electron spectra when appropriate. These data point outgreater than the set pass energy of the analyzer. This was the difficulty associated with accurate determination ofaccomplished by linearly varying the voltage across a spectral shapes of the low-energy wings of the distributionsingle-gap electrode system at the entrance of the at low projectile velocities. At higher projectile velocities,analyzer. the two peaks become further separated and therefore the
The experimental chamber was equipped with three sets spectral shapes can be unambiguously determined. Theof mutually perpendicular Helmholtz coils which were data of Fig. 2 for H- also show the presence of the col-used to nullify stray and the earth's magnetic fields in the lisionally induced autodetaching 2s 2)Se state of H-.vicinity of the electron spectrometer. This arrangement Following background subtraction, the data were digitizedpermitted the magnetic field to be reduced to - I mG in preparation for deconvolution with spectrometer reso-over the spectrometer volume. Electron spectra were tak- lution function.en with a standard x-y recorder with abscissa and ordi- Spectra deconvolution. Various methods of deconvolut-nate directions, respectively, driven by signals from a pre- ing experimentally determined data from calculated, as-cision ramp voltage generator and a conventional channel sumed, or measured instrumental resolution functionselectron-multiplier circuit. A typical count rate at the have been described in the literature. The technique is ofpeak of the detachment signal was _ 104 electrons/sec. practical importance in that it, in principle, permits ex-
The observed laboratory energy Eah of an electron with traction of the true shapes and distributions of spectraenergy Ee and mass me ejected collinearly with respect to data and enables, for example, the determination of abso-a moving ion of energy E, and mass M, is given by the lute energy and transformed-velocity distributions. The
* following small-angle kinematical approximation: result of the pro,.-ure is to accentuate structures withinthe data by it creasing peak heights while effectively de-
EI.,hA 2 [meE E +Ee creasing their spectral widths..,, The procedures used to deconvolute data displayed inthis article follow those described previously in Ref. 30
where the positive and negative signs refer, respectively, to but with appropriate modifications. The observedejection of electrons in the direction of or opposite to the electron-detachment rate R (c(] ) at the jth value of themotion of the beam. average electron energy (e1 ) is an integral product of the
The center-of-mass energy E, of autodetachment elec- analyzer resolution function (energy-distribution function)trons ejected in the forward direction from a moving ionbeam of energy E, can be determined relative to the elec-trons which move at or near the projectile velocity by useof expression (I). The resulting relation for Ee is given by
11/2 t/ H + Ar - H + Ar + ee - (2) H- ION ENERGY 204ev
E H-12s2
) AUTO
where Xi and Xe are, respectively, the potentials through DETACHED ELECTRONS
which the electrons moving at or near the velocity of the -' ENERGY
ion beam and the autodetaching electrons are accelerated z -or decelerated upon entrance into the analyzer, and e is k "".
the electronic charge. By using this procedure, based on B ROUND FUNCTON
differences, errors due to contact and surface potentials -are minimized and the need for precise knowledge of thespectrometer constant and energy-scale calibration is 0 0 20 30 40 o0avoided. E, ev.
III. DATA-PROCESSING PROCEDURES FIG. 2. Collisionally indtuced energy spectrum for electronsdetached from 20-keV H ions interacting with Ar. The back-
Background subtraction. The spectral data, taken by ground function shown is typical of those used in data reductionuse of the procedures outlined previously, were subjected for extraction of low-energy projectile-electron-detachment spec-to background subtraction and digitized for deconvolution Ira.
31 ELECTRON-DETACHMENT SPECTROSCOPY OF 20-100-keV ... 2133
" ff ((Ej),E) that corresponds to the same value of the aver- I
age electron energy (e,) and g(e), the monoenergetic de- He- ON ENERGY 80keV
tachment rate at energy e. The observed detachment rate AUTO-DETACHED NO TARGET
can, therefore, be represented according to the following ELECTRONSt cM ENERGY 19 8 eV)
expression:R- R((cjE/ , g i ')J't( cj),')dc' , (3)
- where the integration extends over all e'. The solution toEq. (3) is obtained iteratively. Estimated values of gc) WMO77eu-\
are introduced into Eq. (3) along with the measured or "analytical energy-distribution function J'((E(),-) and thevalue R ((E I is calculated for every value of (cj) forwhich detachment rates have been measured. A weight- z
ing factor w) is then defined as the ratio of experimentally Zdetermined detachment rates R,,, to calculated detach- 0
ment rates R,.,, from the following expression: m/--BACK GROUND
()( (C 1 ))(R\.,p ((ej ) /R 1ak( (e )~ I (41 BCGRUO.
with the condition that ,(() = I wheneverR,.,&.(4j )=0. The weighting factor w is then used todefine a new g,,, () through the expression
gne,, (f)= goid (f)I &)l (ej ) )fl fEj ), C) f( (,j ), C) . (5)I J I I I -
Equation 3) is repeatedly solved until the residuals de- 3E LECo 60 ENE GY0.,ELECTRON ENERGY (eV)
fined by FIG. 3. Energy spectrum for electrons autodetached from"[Rexi( (e) s- R i ( (e1 ) 1]- -. minimum 16) 80-keV He- ions.
I%
reach minimum values. When this condition is satisfied,the weighting factor (o( (eI ) ) - I and the function g(e) targets. Accumulated data were subjected to backgrounddoes not change with successive iterations. The function subtraction, digitized, and deconvoluted from the spec-g(E) represents very nearly the absolute electron- trometer resolution function. A typical autodetachment-detc reet vergnerly the- electron energy spectrum resulting from the decay of thedetachment rate at energy c.4o
The energy-distribution function f ((E ),e) is normal- P' states of He- was used as the spectrometer resolutionfunction for the deconvolution of all ELC data. Choice
ized to unity for all energies C.j or of this spectral distribution for the spectrometer resolu-
SJf(( ),e')de'= 1 . (7) tion function is based on the near monochromaticity ofthe autodetaching electrons from He, the energy in-
The spectrometer resolution function. The resultant de- dependence of the FWHM r of the observed spectra (seetachment rate gle) depends, of course, on a judicious Fig. 4) and the fact that the spectra were determinedchoice of the spectrometer resolution function ft (e1 ),e). under precisely the same experimental conditions as thoseIdeally, f( ( 1 ),c) should be determined experimentally by existing during the measurement of the ELC spectra.measuring the apparent energy distribution from amonoenergetic source of electrons under conditions identi-cal to those used to obtain R ((ej)) (defining apertures, 2.0
energy, etc.), the data of interest. The resolution functionused in the deconvoluted spectra displayed in this paperwas determined by measurement of the energy-distribution function from the autodetaching state of
4p Qm1.2)'pHe- (= 19.74 eV)-a typical spectrum of which is 2displayed in Fig. 3. The FWHM r of the autodetaching ,.0electron was found to be essentially independent of He- 0.•
projectile energy as shown in Fig. 4 and, therefore, the a 0.,
same function f( (cj ),c) was used for deconvoluting all 0.4
H -- spectral data. 0.2
0.0
IV. RESULTS AND DISCUSSIONS a 20 40 so so 100 o10 140 160
EtlkeV)
Electron loss to the continuum (ELC) spectral data for FIG. 4. Variation of the FWHM r of the autodetachmentelectrons emitted in the forward direction were measured electron spectra from the 4PJ' state of He with projectile energyfor 20-100-keV H- projectiles interacting with thin Ar E,
* 2134 G. D. ALTON, R. N. COMPTON, AND D. J. PEGG 31
FIG. 5. ExperimentalY determined energ and transformed-velocity spectra for electrons detached from 20- 100-keV H- projec-tile% interacting itth t hin Ar target%. Note the appearance of structure in the low%-energy side of the l00-keV spectrum.
*All experimental ELC spectral data displayed in this re- H- electron -detachment spectral data. Collisionally in-* port were normalized to unity at the maxima prior to duced ELC energy and velocity spectral data derived from
deconvsolution. The comparison deco.nvoluted spectra H_ interactions with thin AT targets are displayed in* were then computed relative to the ricrmalized experimen- Figs. 5 and 6. Although there are no pronounced struc-
tal values. Solid lines passed through the experimental tures in these data, several aspects are worthy of corn-points of the spectra are added for clarity and are not ment. First of all, we note a progressive increase in theleast-squares fits to the data. base widths of the low- and high-energy wings of these
FIG. 6. Deconvoluted (unfolded) energy and tratisformed-s'elocity spectra for electrons detached from 20-- lO0-keV H -projectiles
interacting with thin Ar targets. Note the appearance of structure on the low-energy side of the l00-keV spectrum.
31 ELECTRON-DETACHMENT SPECTROSCOPY OF 20-100-keV ... 2135 __
0.14 , I l ] I
0.12 ]
0.10 .
0
-. 0.08 - -- - ' -0 -
0.060.04 - - r (a.u.) - 3/2 00 V, (a.u.)
0.04 _ .-- __
0.02
0.00 I I I
0.8 1.0 1.2 1.4 1.6 1.8 2.0 2.2 2.4 2.6 2.8
ELECTRON VELOCITY V. (a.u.)
FIG. 7. Variation of the FWHM F with electron velocity v, of the electron-detachment spectra of electrons ejected from 20-100- - -
keV i projectiles interacting with thin Ar targets. The theoretical curve is that predicted from first Born calculations usingprtjectile-centered Coulomb final-state wave functions.
spectra as the projectile energy increases. In addition, the aspects concerning the collisionally induced electron lossspectra exhibit rather striking asymmetric cusplike to the continuum (ELC) process for H- projectiles. Thecharacteristics in spite of the dominance of single-loss significant findings of these investigations are summa-processes in this velocity regime.' (Single-loss cross sec- rized below.tions are approximately an order of magnitude greater The electron energy and velocity spectra for H- col-than those for double loss. While the ELC spectrum at lisions with Ar targets are observed to be surprisingly20 keV is symmetric, the higher-energy spectra are cusplike in shape even though single electron-loss process-skewed slightly asymmetrically toward the higher electron es are known to dominate. 31 (The cusplike character isenergies. We also note a gradual appearance of structure generally construed to be associated with attractivein the low-energy wings of the 100-keV spectrum dis- Coulomb final-state interactions as predicted by first-placed with respect to the center of the ELC spectrum by order Born approximations.) The ELC spectra for H- .1
13.5 eV-the position of the Ar target-dependent in- are symmetric at lower ion energies and progressively be-terference phenomena observed first by Menendez and come skewed for higher ion energies. The target-Duncan. ' 2t - The lack of such structure at lower energies dependent interference phenomenon discovered by Menen-is also commensurate with observed trends toward de- dez and Duncan - t 2 is depressed for energies below 100creasing amplitudes of this phenomena with decreasing keV where it becomes faintly observable. The FWHM rprojectile energy. for H- interacting with Ar is observed to increase in an
The dependences of the FWHM on electron velocity are approximately linear fashion with electron velocity.displayed in Fig. 7. The solid line represents a linear Analogous measurements of collisionally inducedleast-squares fit to the data, while the dotted line shows projectile-electron-loss processes for other negative ionsthe theoretically predicted dependence of F for attractive are in progress. .-
Coulomb final-state interactions with F= - ,vi. The ex-pression is valid in the asymptotic limit of v, >->v,, where ACKNOWLEDGMENTSv,, is the velocity of the electron before detachment. AK WE MN
V. SUMMARY AND CONCLUSIONS This research was sponsored by the U.S. Department of
Energy, Division of Basic Energy Sciences, under Con-
Experimental investigations have been made of electron tract No. DE-AC05-840R21400 with Martin Maiettaloss to the continuum (ELC) of forward-ejected electrons Energy Systems, Inc., and the U.S. Department of thecollisionally detached from 20-100-keV H- projectiles Navy, Office of Naval Research, Task 393-071. The au-during interactions with thin Ar targets. This energy thors gratefully acknowledge the assistance of Mr. Kirbyrange is considerably lower than has been previously in- Burkholder in reducing the data and Mr. Verner Ander-vestigated. The investigations reveal several interesting son for the loan of the deconvolution computer program.
2136 G. D. ALTON, R. N. COMPTON, AND D. J. PEGG 31
'I. A. Sellin. J. Ph%. IParis% Coloq. Suppl. No. 2,40, Cl (1978. 15F. Drepper and J. S. Briggs, J. Phys. B 9, 2063 (1976).2M. Breing, S. B. Elston, S. Huldt, L. Liljeby, C. R. Vane, S. D. 1j. S. Briggs and F. Drepper, J. Phys. B I1, 4033 (1978).
Berry, G. A. Glass, M. Schauer, 1. A. Sel.in, G. D. Alton, S. 17 M. W. Lucas and R. Strong, Bull. Am. Phys. Soc. 23, 1087 7
Datz, S. Overbury, R. Laubert, and M. Suter. Phys. Rev. A (1978).25, 3015 (1982). 18J. S. Briggs and M. H. Day, J. Phys. B 13, 4797 (1980).
'M. E. Rudd, C. A. Sautter, and C. L. Bailey, Phys. Rev. A 3. 19M. H. Day, J. Phys. B 13, L65 (1980).1635 (19661. 20M. G. Menendez and M. M. Duncan, Phys. Rev. Lett. 40,
4G. B. Crooks and M. E. Rudd, Phys. Re%. Lett. 25, 1599 1642 (1978); Phys. Rev. A 19,49 (1979); 23, 1085 (1981).1970). 2 1M. G. Menendez and M. M. Duncan, Phys. Rev. A 16, 1799
'K. G. Harrison and M. W. Lucas, Phys. Lett. 33A, 142 170): 11977).35A, 402 (1971). 22M. G. Menendez and M. M. Duncan, Phys. Rev. A 20, 2327
l 1W. J. B. Oldham, Jr.. Phys. Re,. A 140, 1477 (19651: 161, 1 (1979).(19671. 2_M. R. Franz, L. A. Wright, and T. C. Genoni, Phys. Re%. A
'T. F. M. Bonsen and L. Vriens, Physica (Utrecht( 47. 307 24, 1135(1981).11970). 24 M. H. Day, Phys. Rev. A 26. 1260(19821.
'A. Salin, J. Phys. B 2, 631 11969): 2, 1225 ( 1969); 5, 979(1972). 25N Maleki and 1. Macek, Phys. Re%. A 26, 3198 ([982)-.'I. Macek, Phys. Rev. A I, 235 (1970). 26V. H. Ponce and R. A. Baragiola, J. Phys. B 17, 2467 (1984). -('K. Dettmann, K. G. Harrison, and M. W. Lucas. J. Phys. B 7, 27C. R. Garibotti and J. E. Miraglia, J. Phys. B 14, 863 (1981).
269 (1974). 28J. Burgddrfer, Phys. Rev. Lett. 51, 374 (1983).1'C. R. Garibotti and J. E. Miraglia, Phys. Rev. A 21, 572 2'(G. D. Alton, R. N. Compton, and D. J. Pegg, Phys. Rev. A
(1980). 28, 1405 (1983); R. N. Compton, G. D. Alton, and D. J. Pegg,'2C. R. Vane, 1. A. Seflin, S. B. Elston, M. Suter, R. S. Thoe, G. J. Phys. B 13, L651 (1980).
D. Alton, S. D. Berry, and G. A. Glass, Phys. Rev. Lett. 43, 3()L. G. Christorophou, D. L. McCorkle. and V. E. Anderson, J.1388 (1979). Phys. B 4, 1163 (1971).
I'M. Rodbro and F. D. Andersen, J. Phys. B 12, 2883 (1979). 31J. Heinemeier, P. Hvelplund, and F. R. Simpson, J. Phys. B 8.14R. Shakeshaft and L. Spruch, Rev. Mod. Phys. 51, 369 (1979); 1880(1975).
Phys. Re%. Lett, 41. 1037 (1978).
=S °
Volt % 54. NUMBER 5 PHYSICAL REVIEW LETTERS 4FEBRUARY 1985
Photoelectron Angular Distributions for Near-Threshold Two-PhotonIonization of Cesium and Rubidium Atoms
Adila DodhyPhisics Department. Auburn University. Auburn. Alabama 36849. and Oak Ridge National LaboratorY. Oak Ridge.
Tennessee 37831
and
R. N. Compton and J. A. D. StockdaleChemical Physics Section. Health and Safe.y Research Division, Oak Ridge National Laborator, Oak Ridge, Tennessee 37831
(Received 1 October 1984) -
Photoelectron angular distributions have been measured for nonresonant two-photon ionizationof cesium and rubidium atoms just above the ionization threshold. The photoelectron energiesranged from 25 to 100 meV. The results are compared with theoretical estimates based on nonrela- . "tivistic atomic wave functions. Initial results are also presented for above-threshold ionization incesium.
new insights into various problems of atomic structure Ray, DCR-lI, PDL-I) was crossed orthogonally by aand dynamics.l- 3 Perhaps the most powerful approach thermal alkali-metal beam. The dye-laser pulse dura- -.is the measurement of differential cross sections,4t-t tion was 5 ns and the bandwidth was 0.02 nm. Thewhere the angular distributions of the photoejected laser was focused to a power density of 108 W/cm 2 byelectrons provide data not only on the magnitudes of a 35-mm lens. The power density was an order of - -
the transition amplitudes but also on their relative magnitude greater when electrons from above-phases. In addition to providing information about the threshold ionization were studied. The plane of polari-scattering phase, thus complementing single-photon zation of the laser was rotated by a double-Fresnelstudies," such measurements also test our theoretical rhomb. Photoelectrons emerging perpendicular to theunderstanding of high-order bound-free transitions in- propagation vector of the laser beam and within ± 2*volving sums over virtual intermediate states.t 2 were energy analyzed by a 160 ° spherical-sector elec-
Studies of photoelectron angular distributions for trostatic energy analyzer. They were then detected byalkali-metal atoms have been limited to cases ofresonantly enhanced MPI 44 ' '" ° or higher-order non-resonant processes.' 2'' 3 In this paper, we report pho-toelectron angular distributions for nonresonant two-photon ionization of cesium and rubidium atoms ,Rbwhere the photoejected electron has an energy in the Crange - 25-100 meV. Figure 1 shows the ionization 30scheme for both alkali metals. The first photon liesbetween the 6p and 7p states for cesium and between 7- 6p 2 P
the 5pand 6pstates for rubidium. 7p pOur measurements are novel in two respects. First, 2 20
we have studied photoelectron angular distributions in "a region very close to the ionization threshold. This is .p 2 pdifficult experimentally because of the very low energy 6p 2 5 .
of the photoelectrons under consideration. Second, 'owe report photoelectron angular distributions for 'above-threshold ionization of cesium and compare . .
them with theoretical predictions. Such processes "' -2-",
have been observed by others in xenon14 -18 and cesi- 06S 2S 5s "
um 3 but only in higher order (order 5 in cesium and6 in xenon).
Details of the experimental apparatus have been FIG. 1. Energy-level diagram showing the excitationdescribed recently' ° in conjunction with resonantly scheme leading to ionization for nonresonant two-photon -. ,
enhanced MPI. Briefly, the output from a Nd:YAIG ionization of (a) cesium atoms and (b) rubidium atoms.
L V()I Nil 54. Nt Wt K 5 HIYSIC.AL. RELViI, EW I IRS 411 1h I \R '
a dual-channelplate charged-particle detector, and theamplified signal was fed into a gated boxcar integrator X -632 5 .... 62"
(Princeton Applied Research, Model 162). Photoelec-tron angular distributions were obtained from a recordof tht relative photoelectron intensity as a function of -the angle, 0, between the polarization of the laser and -
the fixed direction of detection of the photoelectrons. .The laser-alkali-metal-atom interaction volume wascarefully shieided from external electric and magnetic ,fields in order to detect the ultraslow photoelectrons. __"""""_
All critical surfaces were also coated with colloidal X -6'45_--X "2-0"-graphite in order to reduce surface potentials and elec- U I ,tron reflection.
Cross sections for photoelectron angular distribu-tions for two-photon ionization may be derived by useof time-dependent perturbation theory. 19,20 Theresulting differential cross section, r(f)' 21 , is given by - . •
I (r(O) ( 2111 = ('(21( 1 + o 2) OS +j342) COS4fl, (),9 ....
wh (110 V/2 VO 2where C. 2' is a normalization constant. The coeffi- ANGLE ons) ANGLE,/ rodons.
cients j3,(2 are ratios of linear combinations ofsecond-order radial matrix elements, r/2 , and cosine (b)functions of the difference in phase shifts between the x5e9o,nm X.586 0 nm
* allowed I=0 and 1=2 continuum waves. The radial .matrix elements are given by (2)
where Eg, and E,, are the single-particle energies, w S-
is the frequency of the radiation field, k is the wave 4_j
number of the photoelectron, and (nllrlnl') are radial , 1dipole matrix elements. The radial wave functions for X 5840 , X 582.0nm I"the ground states were generated in the nonrelativistic I.'Hartree-Fock approximation. The sum over the virtu- -'
/ J-
al intermediate states, Inp), found in Eq. (2) can bedetermined by any appropriate method. For the pur- '/ , -poses of comparison with the experimental data, we i"...will use the results of two independent calculations:./(I ) an analytic-expansion method based on a Sturmian .basis set, 21- 23 and (2) an inhomogeneous-differential-equation method 24 - 2' based on a Hartree-Fock poten- 0 ,7/2 W 0 /2 ,tial. ANGLES (rodws)
The experimental and theoretical photoelectron an- FIG. 2. Photoelectron angular distributions for non-gular distributions are shown in Fig. 2. The two- resonant two-photon ionization of (a) cesium atoms and (b)photon ionization threshold corresponds to a laser ex- rubidium atoms. The error bars are three times the size ofcitation wavelength of 636.8 nm for cesium and 593.6 the dot. The solid lines are the theoretical calculations usingnm for rubidium. The solid line through the experi- the Sturmian-basis-set method and the dashed curves aremental data points is the theoretical calculation using from the Hartree-Fock procedure.the Sturmian-basis-set method, 21 while the dashedcure is from the tlartree-Fock procedure.24 The ex-perimental and theoretical intensities are normalized determined by least-squares fitting the data with Eq.to each other at , = 0' Good qualitative agreement is (I). For cesium, the Sturmian procedure gives results
, seen between the experimental measurements and which are actually in quite good quantitative agree-theoretical calculations for both atoms The experi- ment with the experimental results. In comparing the
- mental and theoretical values for the 11 coefficients are photoelectron angular distributions for the two atoms.shown in Table I The experimental values were we note that both experiment and theory find that the
ratio of the intensity at 0 = 90' to that at 0 -0* is larger for any nonresonant multiphoton process involving anin rubidium than in cesium for approximately the odd number of photons. It is important to note thatsame value of photoelectron energy above the two- Leuchs and Smith 7 have also observed a nonzero in-photon ionization threshold. We also note that in both tensity at 0 = 7r/2 for five-photon nonresonant ioniza-atoms, this ratio decreases as the laser wavelength is tion of sodium. In our case, the nearest resonant in-decreased. The deviation of the data from the theory termediate state (5d 2D) is over 1000 cmi away fromfor photon energies nearest the threshold may be due the first photon so that detuning far exceeds the spin-to the difficulties with detection of ultraslow electrons orbit coupling and the laser power is far too weak to al-or possibly due to neglect of electron correlation and low for level shifting in and out of resonance. In addi-relativistic effects in the calculations. tion, the polarization, P, of the laser was measured to
For the three-photon ionization of cesium, involving be 0.99 at the laser focus. It is possible that space-above-threshold ionization, a third photon is absorbed charge and/or nonzero background effects coupledin the continuum. The differential cross section,r (0)13? is given byi 9
o. (9)(3) = C 3 1(cos 2g +,3 ) cos40 + 3) cos 60), I - "
(3)where Ct3) is a normalization constant. The coeffi- .[ .cients, p 3), are ratios of linear combinations of third- " ,order matrix elements, rt and cosine functions of ..-
the difference in phase shifts between I= 1 and I=3 ,'"{continuum waves. For comparison with experiment
we will again use the results of calculations which em- .ployed the Sturmian basis set21 and Hartree-Fock pro- , I-:!'! .cedures24 in calculating the summation over inter- -.
mediate virtual states. J 1 0The experimental and theoretical angular distribu- \ / .
tions obtained at a laser wavelength of X = 633.66 nm oare shown in Fig. 3. Both theoretical methods gave * ...essentially identical results. The intensities are nor- / ,malized at 0=00. At 0 =00 the photoelectron peak *
corresponding to the three-photon above-threshold 0ionization was - 1% of that for the two-photon pro-cess. Four- and five-photon peaks were not detected. ANGLE 9 tradions)
If present, their intensity is less than one-tenth that of F 3 h lr n r io r vFIG. 3. Photoelectron angular distribution for above- . ..'the three-photon peak. threshold ionization in cesium at a laser excitation wave-
The measured angular distribition for the above- length of 633.66 nm. The dots are the experimental values. "threshold ionization does not compare well with the The solid line is the theoretical calculation using the .theoretical calculations. The major discrepancy ap- Sturmian-basis-set method and the dashed curve is from thepears at 0 = r/ 2 where theory predicts zero intensity Hartree-Fock procedure. -.
Voitl LiW 54, Nt W~it R 5 PHYSICAL REVIEW LETTERS 4 1- iit %R) IQN;
with the low signal level could partly account lor the 611 Kaminski, J. Ke.ssler, and K J Kolkith. Ph.~ Re%.nonzero intensity at 0 7 r/2. We note that the theory Lett. 45. 1161 (1980)
preict prmarly co2~idistribution whereas the ex- 7 .Leuchs and S. J. Smith. J. Ph\ s. B 15. 1051 (1982).perimental distribution contains contributions from 8D. Feldmann and K. Welge. J. Ph~ B 15. 16 51 (19821,higher-order terms. Further experimental and theo- 9R. Hippler. 11.-J flumpert, 11. Schwier. S. Jet,,ke. and
11-0. Lut,. J. Phys. B 16. L713 (1983)retical studies are in progress. 1R .Cmtn .A ) tcdlC1 oprp ~ W wish N. Cthpton P. A.bop o [). Skd.e Tan of'CoprWe wsh o thnk . Lmbroouls ad X.Tan of X. Tang, and P. Lambropoulos. Ph. Re%. A 30. 1766the University of Southern California and the Univer- (1984).sity of Crete. and M. S. Pindlzola of Auburn University 11S T. Manson and A. F. Starace. Re%. Mod. Ph. 54. 389for numerous helpful discussions and permission to in- (1982).dlude the theoretical calculations. This research was 12j. Morellec. D. Normand, and G. Petite, in .4dtianc's in -
sponsored jointly by the Office of Health and Environ- Atomic and Volecular Phlwisics, edited by D). R. Batesmental Research, U. S. Department of Energy, under (Academic, New York, 1982), Vol. 18. pp. 97-164.Contract No. DE-AC05-840R2 1400. with Martin 13G. Petite. F. Fabre, P. Agostini. M. Crance, and M. Ay-Marietta Energy Systems Inc., and the office of Naval mar. Phys. Rev. A 29. 2677 (1984).Research, under Interagency Agreement No. N000 14- 14P. Agostini, F. Fabre. G. Mainfray. G. Petite. and N. K.85-F-001 1. Rahman, Phys. Rev. Lett. 42, 1127 (1979l.15P. Kruit, J. Kimman, and M. J. van der Wiel. J. Phys. B
14. L597 (1981).16F. Fabre. P. Agostini. G. Petite, and M. Clement. J.Phys. B 14, L677 (1981).
[1P. Kruit. If. G. Muller. J. Kimman. and M. J. %an der1P. Lambropoulos, in Adiaces inAomic and Mlolecular Wiel, J.Phys. B 16, 2359 (1983).Phyics edtedby . . Btes(AcdemcNew York. 1976). 18P Kruit, J. Kimman. H. G. Muller. and M. J. van der
Vo.1,p.8-6.WePv.Rev. A 28, 248 (1983).S2MlihtnPoess dtdb .H Eberly and P. Lam- 19P. Lambropoulos, Phys. Rev. Lett. 28. 585 (1972).
bropoulos(Wiley, Nw York, 978). 20J MizunoJ. Py.B6 1 17)3 Mfultiphoton Ionization of Atoms, edited by S. L. Chin and 21 X. Tang and P. Lambropoulos. to be published.P.Lambropoulos (Academic, New York, 1984). 22S. Klarst'eld and A. Maquet. J. Phys. B 12. L553 (1979)."J. A. Duncanson, Jr., M. P. Strand, A. Lindgrard. and 23E. Karule. J. Phys. B 11, 441 (1978).
R.S. Berry. Phys. Rev. Lett. 37. 987 (1976). 24M. S. Pindlzola, private communication.5.1 C. Hansen, J. A. Duncanson, Jr.. R. L. Chtien, and ISM. S. Pindzola. Phys. Rev. A 17. 1021 (1978).0
R.S. Berry. Phys. Rev. A 21. 222 (1980). 26M. Aymar and M. Crance, J. Phys. B 14. 3583 (1981).
425
-. W W W W W_
Vy ecc*PtOeIC4 01 this artile, th
0Ublijher Of recipient acknowledgesthe 04I. Goviernrnant'8 righ
t torstot n a nonsX"iuivl. rOyeVft-frN
license in endI to &"Y Coipyrigjht
C@Vfron the article. ,
THE PRODUCTION OF Ca THROUGH SEQUENTIAL CHARGE EXCHANGE WITH Li VAPOR*
G. D. Alton, T. J. Kvale,* and R. N. Compton**
Oak Ridge National LaboratoryOak Ridge, Tennessee 37831
D. J. Pegg** and J. S. Thompson***
University of TennesseeKnoxville, Tennessee 37916
The efficiencies for producing Ca- through sequential charge exchange
between Ca+ projectiles and Li vapor have been measured for the first time.
The measurements were made over a projectile energy range of 20 to 100 keV
under experimental conditions commensurate with tandem accelerator applica-
tions. Production efficiencies in terms of negative ion charge state frac-
tions versus projectile energy, Li vapor cell temperature and Li target
density were determined for negative ion formation within a detection win-
dow a of ± 11 mrad. Within this range of angular divergence, the effi-
ciency for production is observed to be an increasing function of the
projectile energy even at the highest energies used during the measurements
(100 keV). The results of these measurements as well as a description of
the experimental apparatus and techniques utilized are presented.
*Research sponsored by the Division of Basic Energy Sciences, U.S.
Department of Energy, under contract DE-ACO5-840R21400 with Martin MariettaEnergy Systems, Inc. and DE-ASO5-83ER13097 with the University of Tennesseeand the U.S. Department of the Navy, Office of Naval Research, TASK393-071.
*Postdoctoral Fellow in the Postgraduate Training Program administeredby Oak Ridge Associated Universities, Oak Ridge, TN.
**University of Tennesse, Knoxville, TN and Oak Ridge NationalLaboratory, Oak Ridge, TN.
***Graduate student at the University of Tennessee, Knoxville, TN.
-2-
1. Introduction
The tandem accelerator concept is predicated on the existence of nega-
* tive ions. Thus, techniques for formation of intense negative ion beams
for a wide variety of elements are of considerable importance to tandem
accelerator based atomic and nuclear physics research programs. In addi-
tion to these traditional applications, an increasing number of other
applications have also evolved over the past few years.E1,21
Negative ion formation through sequential charge exchange interactions
between an initially positive energetic ion and a suitably chosen exchange
vapor offers perhaps the most universal and efficient means of producing
negative ions known to date. Production efficencies depend primarily on
ion energy, the electron affinity of the element under consideration, and
the electron binding energy and density of the exchange vapor. The effi-
ciency of the exchange process depends strongly on the choice of the
projectile-target combination, with highest efficiencies occurring for com-
binations with minimum energy defect in both of the electron capture pro-
cesses.
In general , negative ion formation, for a projectile y and an exchange
vapor x, takes place according to the following spin conserving interac-
t ions.
+ x -X yG + X
0O* + x+ Y*- + x+(1
.y
-3-
where the asterisk indicates the possibility of an excited state of the
atom or ion.
The sequential electron capture process was first discovered experimen-
tally by Donnally and Thoeming[3] in the formation of He- in Cs vapor.
Subsequently, several other investigations have been made of the efficiency
and energy dependence of charge exchange production of H- in Na[4] and Cs
vapors[5] and He- in Li, Na and Mg,[6] K,[7] Rb[8 ] and Cs[9] vapors. In recent
years, a series of systematic investigations has been reported for the pro-
duction of several negative ions in Na and Mg exchange vapors with effi-
ciencies ranging from 0.4% to greater than 90% depending on the electron
affinity of the particular element.[10,11] These investigations also include
production efficiency measurements for Be- in Na(3%)E1[] and Mg(1.751010
vapors and Ca- in Na (0 .4%)[11] vapor at a single energy (50 keV). These
Group IIA elements have negative electron affinities and are known or
believed to exist in doubly excited, metastably bound states. (Recent
experimental evidence has confirmed the metastability of Be-.)[1 2] Since the
formation of Ca- through charge exchange is the subject of the present
study, the result for Ca- in reference 11 is of particular relevance. The
Group IIA elements are examples of a class of elements that are well suited
for charge exchange negative ion production since they are difficult to
produce by other means - provided, of course, that they can be formed in
states which live long enough to be of practical value.